Whether or not a high level narrative frame is imposed over game play, most game play in interactive 3D games occurs at a level where the very high number of possible combinations of actions, positions and interactions makes it completely inappropriate to map out branching interaction possibilities at the level of the design of game moves. Interactions and their consequences between paired types of actions can be specified, using, for example, payoff matrices. But in the simulation of a continuous game space where players can choose moves for interacting with other player characters or NPCs, the details of interaction are far too great to consider other than at the level of the pair-wise interactions of moves that are highly combinable in sequences. This amounts to the design of game characters encapsulating interaction potential in their available potential moves, constituting a character- (and game object-) level object-oriented approach to interaction design.
From a story perspective, this amounts to the adoption of object-oriented story construction methodologies (see ), in which game objects (including characters) encapsulate their own potential for the construction of interesting story material. What represents interesting story material depends upon the play experience preferences of the player. As noted above, game moves represent a performance repertoire of in-game player character performance primitives. Combat-oriented games, for example, provide player moves for actions such as moving the player character within the game world, exchanging, selecting and deselecting weapons, aiming and choosing opponents, and actually striking opponents (depending upon the game; for MMORPGs, where time delays exclude fine timing, the move is to initiate combat, with each blow typically being executed autonomously). Combat moves fit into a conception of interaction conforming strongly with the formal structure of a game. Combat moves are nevertheless performance moves that can also constitute player performance primitives for manifesting the lowest level of interactively selected detail in the game experience regarded as an unfolding story. Since the character encapsulates it’s movement capacity, this is at least conceptually a variant of object-oriented story construction. For strongly combat-oriented games, however, a small number of types of moves are performed in very high density, that is, repeatedly at very short time intervals and for very long periods of time. This results in the game-intensive form of combat confrontations (each enemy defeated is one game bout won) dominating the player’s experience of play. Performance of the game moves (in patterns or repetitive gameplay gestalts) consumes most or all of the player’s attention, leaving little sense of higher level story or character development. This combat orientation means that levelling and increasing combat capability are the most relevant aspect of character development; any other elements of characterisation or story then fade into the background, and the details of higher level narrative contexts are frequently forgotten (as argued in more detail in Lindley, , ).
The solution to the difficulty of remembering a high level narrative context during highly repetitive game play may lie in good design of the methods by which the narrative context is presented, such as frequent use of cut scenes well integrated into an overall rhythmic pattern of game play. The problem of repetitive game play being more game-like than story-like is of a very different nature. This is, in fact, not a problem for the gamist, and the well designed narrative framing may not be a problem for the gamist/gamer who likes the occasional break (often designed as a reward for achievement in the game), resting for a while in the audience-oriented story role of the viewer of a cut scene. It is, however, a problem for players having the story orientations of performer and immersionist, since repetitive game play (typified by combat games) does not provide a sense of dramatically significant character or story development within the detailed interaction mechanics of play.
Object-oriented storytelling implemented at the character level can use linear or branching narrative (or narrative-like) structures to model character development over time within the control model of a character. This is a weaker sense of narrative than the use of a high level narrative form covering the interactions of many game characters and events. However, it is a way of structuring the development over time of the psyche, beliefs, affective states and behaviours of a character. This can be referred to as the creation of inner narrative models. This concept is closer to the concept of narrative developed by Louchart and Aylet , but provides a more specific advanced plan for character development. Progress along the inner narrative model can be a function of game world and character interactions, thus helping to shape the history of the game world in general but not necessarily leading to any specific overall and high level pattern of development (eg. narrative) at the world level.
A final point about object-oriented storytelling is that it can be applied at the world story level without being encapsulated within character models. An example of this is to use plot controllers that create specific transformations in the world and characters. A specific plot point is represented by a set of conditions marking its activation and a set of rules for transforming world and character states. An example may be a “falling in love” plot point controller, activated by a world action. This could, for example, be triggered by having a character look at a portrait of another character, triggering the plot controller which then implements changes to the character control system, making it easier to move closer to the loved one and harder to move away. A set of plot controllers might or might not have internal dependencies, representing a plan for longer term story development. But an important aspect of this concept is that plot controllers are not necessarily activated but function loosely in the world to make interesting things happen under specific circumstances. This can also be used as a strategy for creating inner narratives from a more object-oriented perspective.
Designing for Performative and Immersive Game Play
As seen above, game play framed by a high level narrative presented by cuts scenes, together with expository material found in in-game texts and NPC dialog, satisfies the audience orientation towards story. What then of the performer and the immersionist? RPGs are the game form most clearly directed towards performance. However, as discussed above, characterisation within an RPG game system is generally accomplished using class and race systems providing little beyond features having instrumental effects (eg. combat, movement and magic) for outlining characters, and this is done at the level of character types. Hence the in-game language of characterisation points directly at game mechanics designed for the gamist player, grounding character in repetitive game play rather than dramatic depth and development. Live-action and table-top RPGs benefit from their live performance contexts in giving players the space to imaginatively elaborate characters; indeed, over time, the mechanised rule sets of a campaign fall into the background while the ongoing imaginative improvisation of character and story form the essence of the play experience (see ). Character performance and character immersion are the great strengths of these game forms. Contemporary computer games mechanise game rule systems but do not provide the same scope for imaginative elaboration of character and story by players.
Providing support for performative and immersive story orientations in computer games is therefore the major ongoing challenge for game technology development in support of immersionist and performer game play. Beyond audience-oriented narrative techniques, game play oriented towards dramatic content requires a more object-oriented approach in which the player character encapsulates character and story development potential, and the specific narratives that emerge over time are a function of the players’ history in a game world.
Role-playing systems based upon classes and races provide some initial mechanisms for this. These systems provide accessible, if shallow, characterisation for players who are not skilled in role-playing. Class and race systems are also very general, functioning for new characters who appear anywhere within a game world, with no personal history and no preformed personality. Deeper, richer and more interesting characters require more specific characterisation. This amounts to having a history with its consequences, being associated with specific places and within a complex social structure. This kind of history can develop over time within a massively multiplayer on-line role-playing game (MMORPG), but a new character enters the world like a baby with no family. It really is not a character at all, but a bunch of instrumental capacities. A player who stays within the world for a long time can develop their character into a deeper character having a history and a social network. But the history of such a character is written within a world of endlessly duplicated monsters and objects that respawn or are generically substituted when killed or taken. The character’s history of actions leaves few permanent traces. And the social networks developed over time are defined primarily by players using free text chat facilities, so these identities do not form a game world identify so much as an in-player-group identity. So the markers of character are either generic or exist within relatively closed social networks of small player subgroups. Skilled players may manage to perform and/or immerse within the scope of freedoms allowed by the game system. An ongoing challenge is to provide system support extending effective realisation of these story orientations for less skilled players, and integrating the uniqueness of well developed characters more effectively into the ongoing history and fabric of the game world.
There is noting intrinsic to the forms of immersive and performative game play described here that renders them incompatible with the use of predefined high level linear or interactive narrative frames. The player who does not also like audience-oriented story content may not like such framing any more than the pure gamer or socialiser. For the player who strongly prefers immersive and performative play to the exclusion of audience-oriented story material, systems based only upon character- and game-object level object-oriented story construction provide a solution. However, the question that then arises is: how is it possible to create an experience over time that has the properties of a well-structured narrative?
A central issue here is that of what constitutes a well-structured narrative. For many immersive and performative players the answer may be: a narrative involving a complex and multidimensional character going through both deep and varied emotional experiences and evolving relationships in a (virtual) life of unfolding complexity and interest. Then an object-oriented construction system is a preferred solution. But the answer from highly conventional narrative writing would be: the three-act restorative structure. This leads to a problem for the latter kind of narrativist: how to develop an object-oriented story construction system that in operation results in the formation of a specific high level narrative structure, such as the three-act restorative structure (or Aristotelean dramatic arc, etc.), without defining the plot in advance? This is (essentially) the problem of emergent narrative.
Emergent computing generally refers to computational processes in which the execution of a large number of very simple rules or processes results in the collective generation of interesting higher level patterns or structures. Emergent narrative in the strong sense is concerned with the emergence of well-defined high level narrative forms from the interaction of smaller scale elements (eg. game characters) in a system that does not contain any representation of that high level form. An example of a system that uses an Aristotelean arch of dramatic tension rising to a climax and then declining is the Façade interactive drama system. In this case parameters on story elements indicate a cumulative drama value, providing a basis for dynamic element selection without prespecifying a specific plot ().
If the target narrative form is not so specific, the concept of narrative emergence is correspondingly weakened. For example, the retrospective narrativisation of a character’s experience, in which the experience is internally transformed into a story (considered by Louchart and Aylet, ) must count as a weak sense of emergent narrative. Perhaps an even weaker sense is the process of the player constructing told stories about their play experiences to relate to others after the event. The general Emergent Narrative Theory of Louchart and Aylet  is based upon no high level structural form, but instead uses the term narrative to refer to what here has been described as object-oriented story construction. However, the use of the term narrative in this way loses the stronger meaning of specific high level narrative forms in the way generally associated with the concept of emergence in computing.
This paper has presented a framework interrelating models of game and story form with player preferences for play style and story orientation. Games are regarded in terms of three levels of temporal design, from simulation at the lowest time scale, through the design of game moves above the simulation level, to the structures of specific narrative patterns at the highest level. High level narrative patterns can be designed as a priori linear or non-linear structures, narrative can be approached as an emergent phenomenon, or a game may omit any strong narrative form. Games having no high level narrative design may nevertheless support performative and immersive game play, this being a matter of detailed design of the game mechanics within an object-oriented paradigm of play-time story construction. At a general level this framework can be applied to table-top and live action role-playing games, as well as to computer games; the nature of the experience can be very different, but the same methods of structuring game mechanics in relation to player styles are still applicable.
 Aarseth E. J. 1997. Cybertext: Perspectives on Ergodic Literature. The Johns Hopkins University Press.
 Bartle R. A. 1996. “Hearts, Clubs, Diamonds and Spades: Players Who Suit MUDs”. http://www.mud.co.uk/richard/hcds.htm.
 Bartle R. A. 2004. Designing Virtual Worlds. New Riders Publishing.
 Chatman S. 1978. Story and Discourse. Cornell University Press.
 Dancyger K. and J. Rush. 1995. Alternative Scriptwriting. Second Edition, Focal Press.
 Eladhari M. 2002. Objektorienterat berättande i berättelsedrivna datorspel (“Object Oriented Story Construction in Story Driven Computer Games”). Masters Thesis. University of Stockholm. http://zerogame.interactiveinstitute.se/papers.htm.
 Eladhari M. 2004. “Interactive Narrative”. Unpublished lecture notes from the course Introduction to Game Analysis and Design. Gotland University, Sweden.
 Eladhari M. and C. A. Lindley. 2004. “Story construction and expressive agents in virtual game worlds”. Other Players Conference. Centre for Computer Game Research, IT University, Copenhagen, Denmark, 6-8 December.
 Frasca G. 1999. “LUDOLOGY MEETS NARRATOLOGY: Similitude and differences between (video)games and narrative”. http://www.ludology.org/articles/ludology.htm.
 Frasca G. 2001. Videogames of the Oppressed: Videogames as a Means for Critical Thinking and Debate. Masters Thesis. Georgia Institute of Technology.
 Juul J. 2001. A Clash between Game and Narrative. Translation of Master’s Thesis written in 1999.
 Juul J. 2003. “The Game, The Player, The World: Looking for a Heart of Gameness”. Proceedings, Level Up Digital Games Research Conference: 30- 45. Copier M. and Raessens J. eds.. 4 – 6 November 2003, Utrecht.
 Kim J. 1998. “The Threefold Model FAQ”.
 Klabbers J. H. G. 2003. “The Gaming Landscape: A Taxonomy for Classifying Games and Simulations”. Proceedings, Level Up Digital Games Research Conference: 54 – 67. Copier M. and Raessens J. Eds.. 4 – 6 November 2003, Utrecht, Netherlands.
 Kücklich J. 2001. “Literary Theory and Computer Games”. Proceedings of the First Conference on Computational Semiotics for Games and New Media (COSIGN). Amsterdam. 10-12 September 2001.
 Lindley C. A. 2001. “A Video Annotation Methodology for Interactive Video Sequence Generation”. Digital Content Creation. Earnshaw R. and Vince J. (eds). Springer, ISBN 1-85233-379-0.
 Lindley C. A. 2003. “Game Taxonomies: A High Level Framework for Game Analysis and Design”. Gamasutra feature article. 3 October 2003. http://www.gamasutra.com/features/20031003/lindley_01.shtml.
 Lindley C. A. 2004. “Ludic Engagement and Immersion as a Generic Paradigm for Human-Computer Interaction Design”. Proceedings of the 3rd International Conference on Entertainment Computing. Technical University Eindhoven, Netherlands. LNCS, Springer-Verlag.
 Louchart S. and R. Aylett. 2003. “The Emergent Narrative Theoretical Investigation”. Proceedings, Narrative in Interactive Learning Environments (NILE-04). Edinburgh, Scotland. http://www.victec.org/docs/Nile04.doc.
 Mackay D. 2001. The Fantasy Role-Playing Game: A New Performing Art. McFarland & Company, Inc., Publishers.
 Mateas M. 2001. “Expressive AI: A hybrid art and science practice”. Leonardo, Journal of the International Society for Arts, Sciences, and Technology, 34 (2), . 147-153. Available on-line at: http://www-2.cs.cmu.edu/~michaelm/publications/Leonardo2001.pdf.
 Mateas M. and A. Stern. 2004. Interactivestory.net. http://www.quvu.net/interactivestory.net/.
 Meadows M. S. 2003. Pause and Effect, the art of interactive narrative. New Riders.
 Murray J. H. 1997. Hamlet on the Holodeck: The Future of Narrative in Cyberspace. Baltimore & London: MIT Press.
 Murtaugh M. 1996. The Automatist Storytelling System. Masters Thesis, MIT Media Lab, http://ic.www.media.mit.edu/groups/ic/icPeople/murtaugh/thesis/index.html.
 Nack F. and A. Parkes. 1997. “The Application of Video Semantics and Theme Representation in Automated Video Editing”. Multimedia Tools and Applications. Zhang, H. ed.. 4(1): 57 - 83.
 Nack F. 1996. AUTEUR: The Application of Video Semantics and Theme Representation for Automated Film Editing. Ph.D. Thesis. Lancaster University, UK.
 Nelson T. H. 1982. Literary Machines. Mindful Press.
 Pohjola M. 2004. “Autonomous Identities: Immersion as a Tool for Exploring, Empowering and Emancipating Identities”. inMontola M. and Stenros J. (eds.). Beyond Role and Play: tools, toys and theory for harnessing the imagination. Ropecon ry, Helsinki, 81-96.
 Propp, V. 1968. Morphology of the Folktale. American Folklore Society and Indiana University.
 Rimmon-Kenan S. 1983. Narrative Fiction: Contemporay Poetics, 2nd edition. Routledge: London and New York, 2002.
 Rollings A. & E. Adams. 2003. Andrew Rollings and Ernest Adams on Game Design. Indianapolis: New Riders.
 Ryan M.-L. 2001. “Beyond Myth and Metaphor – The case of Narrative in Digital Media”. Game Studies 1( 1), July 2001. http://www.gamestudies.org/0101/ryan/.
 Stam R., Burgoyne R., and S. Flitterman-Lewis. 1992. New Vocabularies in Film Semiotics: Structuralism, Post-Structuralism and Beyond. Routledge.
 Vogler C. 1998. The Writer’s Journey: Mythic Structure for Storytellers and Screenwritiers. 2nd edition. Pan Books.
 Wardrip-Fruin N. and P. Harrigan Eds.. 2004. First Person: New Media as Story, Performance, and Game. The MIT Press.
 Yee N. 2002. “Facets: 5 Motivation Factors for Why People Play MMORPG's”.
Craig Lindley is Professor of Game Development at the Institution for Technology, Art and New Media, University of Gotland, Sweden. He has previously worked as research manager, Zero Game Studio, The Interactive Institute, Sweden (2001 to 2003), Chief Scientist, Starlab NV/SA, Brussels, Belgium (2000-2001), and Principle Research Scientist, CSIRO Mathematical and Information Sciences, Australia (1989 to 2000). Craig Lindley began his research career working in the area of artificial intelligence and knowledge base systems, with a particular emphasis upon knowledge engineering methodologies, agent control architectures, AI paradigms, learning and adaptation, behavioural robotics and intelligent interactive multimedia systems. An active interest in experimental filmmaking then led him to apply AI methods to dynamic interactive video synthesis and game systems. His current research interests include methodologies for game and interactive media design, game systems, game semiotics and game form, AI in virtual environments, advanced game engine design, believable agents and characterisation, and emergent and interactive story construction systems.
1Published in Bushoff, Brunhild. ed. 2005. Developing Interactive Narrative Content:
sagas/sagasnet reader. Munich: High Text.
2 Bartle (2004) proposes a third dimension for this scheme and discusses alternative categorizations; this paper focuses on the initial four category system, both for simplicity and due to the added validation and refinement supplied by Yee’s (2002) study.
3 MUD is an acronym for Multi-User Dungeon, a text-based form of shared virtual world.
4 A gestalt may be understood as a configuration or pattern of elements so unified as a whole that it cannot be described merely as a sum of its parts.
5 There may be more than two opponents, and each opponent could either be an individual or a group; opponents may also be synthetic, eg. the NPCs of a computer game.